ES2380789T3 - Access procedure to a communication channel for communications networks - Google Patents

Abstract

A method for accessing a communication channel for a communication network formed by a plurality of nodes designed to transmit and receive subframes by means of the communication channel, including the steps of: defining a succession of frames formed by one and the same first number (N) of time intervals associated to corresponding periods of a reference frame formed by a number of periods equal to the first number (N) of time intervals; and executing selectively, at each time interval and by a first node, one from among the operations of: determining, for each period of the reference frame, a corresponding state indication, and transmitting a subframe of its own including a vector (FI) formed by a number of fields, each containing the state indication determined for the corresponding period; and receiving a possible subframe transmitted by another node, and storing the vector (FI) contained in the subframe received. The step of determining a corresponding state indication includes selecting, alternatively, a first, second, third, or fourth state indication, which are all different from one another.

Description

Procedure for accessing a communication channel for communications networks.

The present invention relates to a method of accessing a communication channel for wireless communication networks.

As is known, given a communications network consisting of a plurality of nodes and given physical means that connect said nodes to each other to define a communication channel with a certain capacity, the so-called "access protocols" are now available. to the channel ", that is, protocols that allow the nodes to communicate with each other through the communication channel, sharing its capacity, even if the nodes move through space.

According to the well-known layer model of the International Standard Organization -Open Systems Interconnection (ISO-OSl model), said channel access protocols belong to the so-called "data link layer", also called "media access control layer (MAC) "or" layer two ". Since they are included in the so-called "MAC layer", the protocols for access to the channel are also called, for short, "MAC protocols".

In more detail, to communicate, the nodes access the communication channel, that is, they transmit signals (commonly electrical or electromagnetic signals) through the communication channel. Such signals transmit information, and the elementary unit of information transmitted at the MAC level by an individual node is generally known as "frame". The channel access protocols conceive the control of the accesses of the nodes with the purpose of optimizing the exploitation of the capacity of the communication channel and reducing the probability of collision, that is, the interference between the signals transmitted by the different nodes.

With regard to wireless communications networks in particular, the so-called “Carrier Sense Multiple Access / Collision Avoidance” (CSMA / CA) protocol is known, for example, which is used, among others, by the IEEE 802 family standards , 11, commonly known as Wi-Fi standards.

The CSMA / CA protocol provides that each node verify before transmitting, the status of the communication channel, that is, verify that the other nodes of the network are not making any transmission, and transmit only in case the communication channel be free On the other hand, in case the channel is busy, that is, if the channel is being used by another node, the node waits for a random period of time or "timeout" to verify the communication channel status again and, in case the communication channel is finally free, transmit.

The CSMA / CA protocol also provides that, after reception by a receiving node of a frame sent by a sending node, the receiving node sends an acknowledgment frame, called "ACK frame".

In the event that a sending node receives the ACK frame, it will consider that the receiving node has correctly received the previously sent frame.

Instead, if the sending node does not receive the ACK frame, it deduces that a collision has occurred. In this case, after a new waiting time, the status of the communication channel is verified and, in case the communication channel is free, the transmission of the previously sent frame is repeated.

The CSMA / CA protocol has achieved considerable success in the field of wireless communications, in particular due to the excellent operation of the communication channel in the case of a network that is substantially free, that is, in the case of a network in that the traffic exchanged, that is, the set of information exchanged by the nodes, is much lower than the capacity of the communication channel. However, in the presence of heavy traffic and, therefore, in case of network congestion, the nodes must endure considerably long waiting times to access the communication channel, that is, transmit; whereby the CSMA / CA protocol is far from scalable. In addition, since no mechanism is reserved for reserving access to the communication channel, it is not possible to guarantee the so-called "quality of service" (QoS).

A possible alternative to the CSMA / CA protocol is that represented by the protocol described in the document "A fivephase reservation protocol (FPRP) for mobile ad hoc networks", "Wireless Networks", January 1, 2001, Kluwer Academic Publishers, The Netherlands , vol. 7, pp. 371-384, by Chenzi Zhu et al. In particular, the protocol proposed in that document provides for stages of assigning time intervals to the nodes of the network. However, despite the fact that in principle this guarantees a certain quality of service, this protocol is optimal in cases where, during the allocation stages, the network topology remains fixed. On the other hand, if during the allocation stages one or more nodes enter or leave the network, there is a statistical deterioration in the performance levels of said protocol.

To overcome the inconveniences described and in particular to guarantee the quality of service depending less

of the behavior (input / output, transmission, etc.) of the nodes of the network, the protocol called "RR-Aloha" has been proposed, described in detail in the document Wireless Networks 10, 359-366, 2004, published in Holland by Kluwer Academic Publishers. The RR-Aloha protocol is also mentioned in the document "Reuse Efficiency of Point-to-Point Connections in Ad Hoc Networks", Global Telecommunications Conference, 2007, Globecom '07, IEEE, Piscataway, NJ, USA, November 1, 2007 , pp. 4494-4499, by L. Campedelli et al., Where it is described with particular reference to point-to-point communications in fixed networks. In addition, the RR-Aloha protocol is also mentioned in the document "ADHOC MAC: New MAC Architecture for Ad Hoc Networks Providing Efficient and Reliable Point-to-Point and Broadcast Services", Wireless Networks, The Journal of Mobile Communication, Computation and Information , Kluwer Academic Publishers, DO, vol. 10, No. 4, July 1, 2004, pp. 359-366 of F. Borgonovo et al.

In particular, the RR-AIoha protocol provides that the nodes of the network have respective MAC addresses and that, in the first instance, they are synchronized (using, for example, the global positioning system, GPS). In addition, the RR-AIoha protocol provides that the communication channel is divided into elementary time intervals, that is, it is one of the so-called "slotted channels".

As shown in greater detail in Figure 1, the RR-AIoha protocol envisages dividing the time into cycles with the period T. Each cycle contains a frame 1, which in turn is divided into a number N of consecutive time intervals 2 of equal duration d, within which network nodes can transmit. In particular, information transmitted in a single time interval is commonly referred to as "subframe 3".

As also shown in Figure 1, the duration d of the time slot 2 is adequate to allow a single subframe 3 to be transmitted in a single time slot 2. In particular, taking into account the geographical area of the network of communications and the maximum propagation times used in the propagation of the signals through said geographical area, the duration d is adequate to allow a sending node to send a single subframe 3 in a single time interval 2, the reception being possible of said single subframe 3 by any other node of the network in said time slot 2.

In addition, the RR-AIoha protocol provides that each subframe 3 comprises a payload part 4, which contains information from layer two, such as a succession number, a source MAC address, a destination MAC address, a Fragment number and a CRC code. Also, the payload part 4 usually contains the possible layer one information associated with the communication channel. Each subframe 3 further comprises an additional part, hereinafter referred to as "frame information vector" (FI) 5, described in detail below. It is also provided that the subframes 3 have a shorter duration than the duration d of the time intervals 2, such that each time interval 2 further comprises a safety time Tg.

Within a frame, each node can transmit in one or more time slots 2. The RR-AIoha protocol works in such a way that it prevents, where possible, two different nodes collide when they transmit in one and the same time interval 2 For practical reasons, the RR-AIoha protocol provides for the association, at each node, of at least one respective transmission time interval of the N time intervals 2 mentioned above, within which the node can transmit.

To determine the associations between the nodes and the respective transmission time intervals in order to prevent collisions, each node maintains the states of the previous N time intervals in memory, at each time interval, each of which can be states, alternatively, "free" or "occupied" by a node. In the following, for short, it will be indicated that the time intervals are free or occupied, taking into account that the terms "free" and "occupied" refer to the states of the time intervals. In addition, it is also said that a node considers that a time interval is busy or free to indicate that the node has in memory a state relative to said time interval, and that said state is "busy" or "free".

When a node accesses a certain time interval, that is to say it transmits a subframe, it adds to said subframe the states of the N-1 time intervals that precede the determined time interval, stored in its memory, in addition to the state of the interval of determined time, identified as busy. In case the time intervals are occupied, the node also adds to the subframe the corresponding identifiers of the nodes that have occupied said time intervals. For example, in each subframe there is an indication relative to the fact that the time interval in which the subframe has been transmitted is occupied, and an identifier of the node that has transmitted the subframe.

In general, given a succession of frames, and therefore of time intervals, it is possible to establish a correspondence between said time intervals and N time intervals of an archetype frame (not shown), hereinafter referred to as "intervals of archetypal time. " In fact, in general, the k-th time interval k of the sequence of time intervals corresponds to an archetype time interval k mod N of the archetype frame; for example, assuming that N equals ten, time interval one and time interval eleven correspond to archetype time interval one. Consequently, add to a subframe the states of the time interval in which the subframe is transmitted and that of the previous N-1 time slots

it means adding the states of each of the archetypal time intervals that form the archetypal plot. In other words, regardless of the number of frames that have actually been transmitted, each subframe contains the states of the N archetype time slots, stored at the time of transmission by the node that has transmitted the subframe.

In particular, to transmit the states of the determined time intervals, the nodes resort to the Fl 5 vectors mentioned above.

As also shown in detail in Figure 1, given a sending node that transmits a subframe in a certain time interval, the vector Fl 5 of said subframe comprises a number of fields 7 equal to the number of time slots present in a frame, which in the case in question it is equal to N. Each field 7 refers to a corresponding time interval (and therefore, a corresponding archetype time interval), which precedes said certain time interval and contains: a temporary identifier of origin (STI), formed by eight bits, which hereafter will also be called STI; a priority identifier consisting of two bits, also called the priority status field (PSF); an occupancy identifier, also called the "busy" identifier, which is made up of one bit and is a function of the state of the corresponding time interval stored by the sending node; and a point-to-point transmission identifier, consisting of a bit and called an FTP identifier.

In more detail, given a field 7, the busy identifier is set to "1" if the sending node considers that the corresponding time interval is busy, that is, it has the busy state stored for the corresponding time interval; otherwise, it is set to "0". In addition, in case the corresponding time interval is occupied, the temporary origin identifier indicates that the node has occupied said time interval. In short, it is also said that a vector Fl indicates that a time interval is occupied by a node in the case where the field of said vector Fl corresponding to said time interval has the busy identifier equal to "1" and the temporary origin indicator that identifies said node.

Compared to MAC addresses (formed by 6 bytes), the use of a temporary source identifier consisting of only eight bits allows reducing the additional information associated with FI vectors, although temporary identifiers can commonly adopt two hundred and fifty-six different values. , thus running the risk of associating nodes other than one and the same temporary identifier of origin.

With respect to the storage of the states of the time intervals by the nodes, the RR-Aloha protocol allows, given a node and the time interval k, for example, that the considered node has in memory, before it is started said time interval k, the states of the N time intervals that preceded the time interval k. In particular, given a nth time interval n of the N mentioned time intervals preceding the time interval k, the state of said time interval n is the state occupied in the following cases:

-

 the considered node has transmitted a subframe in the time interval n, and in the time intervals between the time interval n (excluded) and the time interval k (excluded) the possible subframes received by the node considered confirm the occupation of the time interval n by the considered node; that is, they have FI vectors whose fields corresponding to the time interval n have the busy identifier equal to "1" and the temporary origin indicator indicating the node considered;

-

 the considered node has received in the time interval n a subframe transmitted by another node, and in the time intervals between the time interval n (excluded) and the time interval k (excluded) the possible subframes received by the node considered confirm the occupation of the time interval n by said other node; that is, they present FI vectors whose fields corresponding to the time interval n have the busy identifier equal to "1" and the temporary origin indicator indicated by said other node.

In the rest of the cases, the state of said time interval is free.

The states of the N time intervals that preceded the time interval k refer to the node considered; in other words, assuming that the node considered has determined that the state of a time interval n is occupied, for example, because during the time interval n has received the transmission of a certain node, it is possible that a node different from the node considered, it has determined that the state of the time interval n is free, for example, because, after having moved away from that certain node, it has not received any signal during the time interval n. In short, hereinafter, it will also be said that the time intervals are considered free or occupied by a given node, taking into account that the terms "free" and "occupied" refer to the corresponding states stored by the given node .

On the basis of what has been described so far, the nodes propagate, within the network, the information that they have available for the states of the time intervals, said information being contained precisely within the vectors Fl. In addition, the nodes determine whether they should access a time interval or not according to the Fl vectors contained in the subframes received from other nodes.

In more detail, considering again the k-th time interval k of a given frame and a node that it wishes to transmit but has not yet done so in any time interval of the determined frame, the node determines whether it should transmit or not in said time interval k according to the vectors Fl present in the possible subframes received in the N time intervals that preceded the time interval k.

In particular, the node considers that the time interval k is reserved and therefore does not transmit in said time interval, in the case in which the node has received, in the time intervals between the time interval kN (included ) and the time interval k-1 (included), at least one subframe in which the busy identifier of the field corresponding to the interval kN was equal to "1". Otherwise, the node considers that the time interval k is accessible and accesses it with a certain probability.

In case the node does not really access the time interval k, the described operations are repeated in the time interval k + 1. On the other hand, if the node actually transmits a subframe in the time interval k, then it is selected as its own temporary origin identifier, that is, as a temporary origin identifier that is added to the field of the vector Fl corresponding to the interval of time k, a value between zero and two hundred and fifty-five that has not yet been used by the nodes that have carried out transmissions in the N time intervals preceding the time interval k. As regards, instead, the possible temporary identifiers of origin added to the other fields, these are equal to the same temporary identifiers of origin present in the subframes previously received by the node and, in particular, present in the corresponding fields at the time intervals in which said subframes have been transmitted.

After the transmission, the mentioned node must wait for a frame before it is confirmed that the transmission has been correct. In fact, to establish whether the transmission was successful or not, the node verifies that all the subframes it receives during the N-1 time intervals after the time interval k present Fl vectors in which the busy identifier of the corresponding field at time interval k is equal to "1", and in which the temporary origin identifier indicates the node itself. If so, the node receives confirmation of the correct transmission and can continue transmitting in the time intervals k + iN, i = 1, 2, ..., taking due account of the need to repeat the checks in each transmission mentioned above. Otherwise, the transmission is considered unsuccessful, insofar as a collision is detected in the time interval k, and therefore the operations described are repeated.

The RR-Aloha protocol in concession allows the implementation of a mechanism to reserve access to the communication channel, without the need to resort to a decision-making node, that is, a node that associates, exclusively for each node of the network, a respective time interval in which it must be transmitted. The reservation is actually obtained in a distributed way through the nodes of the network.

The RR-AIoha protocol has proven to be effective in numerous situations and, in particular, in the case of single-cluster networks, that is, in the case of networks formed by various nodes connected to each other, in which the transmission of a node is received by the rest of the nodes. However, there are fields of application in which the use of the RR-AIoha protocol is prone to potential deficiencies, such as in the case of networks whose nodes have high mobility or networks in which frequent inputs and outputs of the nodes, that is, networks in which not all nodes are interconnected. The so-called "ad hoc vehicular networks" (VANET) represent examples of such networks.

A situation in which the RR-AIoha protocol can generate failures is, strictly by way of example, that of a communications network represented in Figure 2 and consisting of seven nodes, designated respectively by n1-n7; in which it is assumed that the number of time intervals N per cycle is equal to ten. In addition, as shown in Figures 3a-3c, which schematically illustrate the states of the N time slots stored respectively by nodes n1 and n3 (Figure 3a), N5 and N6 (Figure 3b) and N7 (Figure 3c), it is assumed that node n1 has transmitted in the zero time interval (it is conventionally assumed that the time intervals are numbered from zero to nine) and that node n5 has transmitted in time interval one. As shown in Figure 2, it is assumed on the other hand that during time interval two both node n2 and node n4 transmit respective subframes. In addition, it is assumed that the subframe transmitted by node n2 will be received only by nodes n1 and n3, due for example to signal attenuation phenomena, and that the subframe transmitted by node n4 will be received only by nodes n5 and n6. It is also assumed that, due to interference between the subframe transmitted by node n2 and the subframe transmitted by node n4, node n7 does not detect any subframe.

In such a situation, nodes n1 and n3 consider that time interval two is occupied by node n2, while nodes n5 and n6 consider that time interval two is occupied by node n4. As regards node n7 instead, it does not detect any transmission; therefore, node n7 interprets that it has not received any subframe in time interval two and, therefore, considers that time interval two is free.

As depicted again in Figures 3a-3c, no node transmits in time intervals three and four. Next, node n6 transmits in time interval five, and the rest of the network nodes receive the subframe transmitted by node n6. In said subframe, the vector Fl indicates, among other things, that time interval two is occupied by node n4. Then, nodes n4, n5, n6 and n7 receive confirmation that the time interval is occupied by node n4.

Instead, nodes n1 and n3, apart from node n2, detect a collision in time interval two and, therefore, according to the RR-AIoha protocol, consider that time interval two is free. This thus generates an asymmetry of the states of the time intervals stored by the nodes of the network. Furthermore, as soon as nodes n1, n2 and n3 mentioned above send a new subframe, incorrect information regarding the status of time interval two is propagated.

Another situation in which the RR-AIoha protocol can generate faults is, as stated above, that represented by networks in which frequent node inputs and outputs occur. In this regard, to simplify the assumption of a frame consisting of three time intervals and, on the other hand, the assumption of a network (not represented) consisting of a first node, which transmits in a first time interval, and a second node, which transmits in a second time interval. In the event that a third node enters the network simultaneously with a third time interval and, therefore, after the first and second nodes have made their transmissions, said third node will not have received the subframes transmitted by the first and The second nodes. In concession, assuming that the third node transmits precisely in the third time interval, the subframe transmitted by it will present a vector Fl in which it is indicated that the first and second time intervals are free. When the first and second nodes receive said subframe, incorrect information is obtained that is subsequently propagated.

The objective of the present invention is to offer a method for accessing a communication channel for communication networks, which at least partially overcomes the drawbacks of the known technique.

According to the present invention, a method for accessing a communication channel, a node for a communications network, a communication network and a software product according to claims 1, 10, 11 and 13, respectively, is disclosed.

In order to achieve a more complete understanding of the present invention, some embodiments thereof will now be described only by way of non-limiting example and with reference to the attached drawings, in which:

-

 Figure 1 is a schematic illustration of a sequence of frames, a time interval, a vector Fl and a field of the vector Fl;

-

 Figure 2 represents an example of a communications network;

-

 Figures 3a-3c are schematic and qualitative illustrations of the information of the frames transmitted by nodes of the communication network represented in Figure 2;

-

 Figure 4 is a qualitative illustration of a data structure and

-

 Figure 5 represents another example of a communications network.

The present procedure for accessing a communication channel represents a type of improved RR-AIoha protocol, in which the nodes, which move in space, detect collisions in each time interval, based on their own detections and of the FI vectors of the subframes transmitted by other nodes.

A possible embodiment of the present process envisages adding, to the fields of the Fl vectors, status indications that are each composed of at least two bits instead of the busy identifier mentioned above (bit busy), as described further ahead. In particular, hereafter, the present procedure is described with reference to the case in which each status indication is composed of two bits, hereinafter referred to as "busy bit" and "CLS bit". In this way, it is possible to detect collisions by comparing different FI vectors and by an explicit collision warning contained in the subframes.

According to the present procedure, at each time interval, each node considered from the nodes of the network may alternatively transmit a subframe or remain waiting for a possible subframe transmitted by another node, it being possible in the latter case that the considered node actually receives a subframe or, for example due to inactivity or collisions, that the considered node receives no subframe. The fact that, during each time interval, the considered node transmits a subframe or, on the contrary, remains waiting for a possible subframe, is determined by the considered node as described below. Hereinafter, however, the present procedure is described with reference to a situation in which the nodes of the

network have already determined in which slots they will transmit or will remain waiting; that is, the subsequent description considers the situation in which the nodes of the network have already sent at least one respective subframe.

In case the node considered receives a subframe, the vector Fl of the received subframe is stored, for example, in a respective memory arranged to contain up to N vectors Fl. In addition, when storing a vector Fl received in a time interval, The node considered associates said vector Fl with the archetypal time interval corresponding to said time interval.

In more detail, at the beginning of a k-th generic time interval k, the considered node deletes from memory the vector Fl associated with the archetype time interval k mod N.

At the end of the time interval k, in case the node considered has not transmitted, the vector Fl of the possible received subframe is stored by associating it with the archetype time interval k mod N. In case the node considered has not transmitted or If no subframe is received, a default Fl vector is stored by associating it with the archetype time interval k mod N. The absence of reception can therefore be considered equivalent to receiving a subframe containing a default Fl vector. In the following, it is assumed, by way of example, that the default vector Fl has zero value fields, including busy and CLS bits, that is, busy bit = 0 and CLS bit = 0.

Assuming, without implying any loss of generality, that the end of the time interval k coincides with the beginning of the time interval k + 1 and, therefore, that the node considered has already deleted from memory the vector Fl associated with the archetype time interval k + 1 mod N, at the end of the time interval k the node considered will present in memory the N-1 vectors Fl of the possible sub-frames received in the time interval k and the previous N2 time intervals, or one or more vectors by default in case of absence of reception.

Since each of the N fields of each stored vector Fl corresponds to a respective archetypal time interval, at the beginning of the time interval k the node considered presents in the memory a matrix data structure, which is constituted, for example, by N-1 rows and N columns, the elements of which are stored FI vector fields that therefore contain status indications, among other things. Hereinafter, indices i and j are used to index, respectively, the rows and columns of the matrix data structure.

By way of example, Figure 4 represents an example of matrix data structure, with respect to a case in which N = 3. In addition, the fields that form the elements of the aforementioned matrix data structure are represented qualitatively with exclusive reference to the status indications contained in the fields and, consequently, to the corresponding busy and CLS bit pairs, that is, without showing the corresponding temporary source identifiers and priority identifiers.

In more detail, Figure 4 refers to the start time of a time interval w as the time at which w mod 3 is equal to zero. In practice, each row i of the matrix data structure is composed of the FI vector stored by the node considered for a period of time h, where w-3 <h <w and h mod 3 = i, where i = 1, 2. In addition, Figure 4 represents in a broken line the part of the matrix data structure that the node considered suppresses at the beginning of the time interval w, that is, precisely the row corresponding to the vector FI stored during the time interval w-3 .

By generalizing what is represented by way of example in Figure 4, at the beginning of each time interval, the node considered has (N-1) x N status indications available, since it has N-1 FI vectors in memory, each of which contains N status indications. Equivalently, for each archetype time interval, the considered node has N-1 status indications available, contained respectively in N-1 fields, each field belonging to a different stored FI vector. In practice, for an archetype time interval with 0 = u = N-1, there are N-1 status indications contained in column j = u of the matrix data structure implemented by the node considered.

Before describing the status indications mentioned above in more detail, it should be noted that, given, given an archetype time interval A, the field that i) belongs to the stored FI vector and associated with said archetype time interval A and ii ) corresponds to said archetype time interval A is called "detection field" and corresponds to archetype time interval A. In other words, with reference to the matrix data structure mentioned above, the detection field of archetype time interval A It consists of the element indexed by i = A, J = A. Furthermore, given a generic node and an archetype time interval, the detection field corresponding to said archetype time interval indicates that the generic node has actually received in the last time interval corresponding to said archetype time interval.

In case the node considered transmits a subframe in the time interval k after deleting from the memory the vector FI associated with the archetype time interval k mod N, N status indications will be added therein

number of fields of the FI vector of the subframe, each status indication referring to one of the N archetype time intervals that make up the archetype frame. Each field of the FI vector of the subframe that has been accessed contains a respective coded status indication, for example, as mentioned above, by the busy bit and the CLS bit. In particular, four different status indications are possible that are associated, respectively, with corresponding pairs of values of the busy bit and the CLS bit of the field and are a function of the Fl vectors stored by the considered node. In other words, the status indications added to the subframe are a function of the status indications present in the stored vectors Fl, as described below.

In more detail, the status indications mentioned above comprise:

-

 a free status indication, associated with the busy bit = 0 and CLS bit = 0;

-

 a busy status indication, associated with the busy bit = 1 pair, CLS bit = 0;

-

 a collision indication, associated with the busy bit = 0, CLS bit = 1 and

-

 an indirect connection indication, associated with the busy bit = 1 pair, CLS bit = 1.

In practice, if the status indications are encoded in the manner described and the default FI vector has zero value fields, the storage of the vector Fl will be equivalent to the storage of a vector Fl, all of whose fields Present the free status indication. However, to generalize, the distinction between free status indication and default indication will be maintained in the future, since in any case the embodiments in which the status indications are coded, for example, by more than two bits, and the default indication differs from the free status indication, the busy status indication, the collision indication and the indirect connection indication, or the embodiments in which the free status indication is not associated with the busy bit = 0, CLS bit = 0.

In operational terms, assuming that the time interval k mentioned above corresponds to an archetype time interval TX1 (k mod N = TX1), the node considered adds, to the field of the vector Fl of the transmitted subframe (more precisely , the one that is about to be transmitted) corresponding to said archetype TX1 time slot, the busy status indication, the temporary origin identifier of the considered node, which hereafter is referred to exactly as "temporary origin identifier", and a priority identifier arranged for attribution to the subframe.

As regards, however, the other fields of the FI vector of the transmitted subframe, the considered node proceeds as described below. Given one of the remaining fields mentioned that corresponds to an archetype B time interval, the considered node takes into account the N-1 fields of the stored Fl vectors corresponding to said archetype B time interval (equivalent to reading column j = B of the matrix data structure implemented), and adds to said field of vector Fl of the transmitted subframe,

1) the free status indication (busy bit = 0, CLS bit = 0), if all the fields considered contain the free status indication or the default indication or if one or more fields considered contain the indirect connection indication, and the remaining fields contain the free status indication or the default indication;

2) the busy status indication (busy bit = 1, CLS bit = 0), if among the N-1 fields considered, the detection field corresponding to the archetype time interval B contains the busy status indication and the temporary identifier of origin contained in this indicates an occupying node, alternatively containing the rest of N-2 fields considered the default indication, the free status indication or the indirect connection indication, or containing the busy status indication and the temporary identifiers of origin indicating the occupant node mentioned above;

3) the collision indication (busy bit = 0, CLS bit = 1), if two or more of the fields considered contain the busy status indication, but contain different temporary identifiers of origin;

4) Indirect connection indication (busy bit = 1, CLS bit = 1), if the detection field corresponding to the archetype time interval B contains the default indication, and one or more of the remaining N-2 fields considered contain the indication of busy status and the same temporary origin identifiers, which refer to a distant node.

In addition, again with reference to the aforementioned field of the vector Fl of the transmitted subframe:

in case 1), the node considered cannot add any temporary source identifier or priority identifier;

in case 2), the considered node also adds the temporary origin identifier of the occupying node, in addition to the

priority identifier contained in the detection field corresponding to the archetype B time interval;

in case 3), the considered node also adds the temporary origin identifier contained in the field that presents the maximum priority identifier of the fields considered to have caused the collision, in addition to said maximum priority identifier, or, in the case that the priority identifiers of the fields that caused the collision are equal, a temporary origin identifier chosen randomly among the temporary origin identifiers contained therein, and

in case 4), the considered node also adds the temporary origin identifier of the far node, and it is also possible, in this case, to also add the corresponding priority identifier to the far node.

As mentioned, the preceding description refers to a situation in which the node considered has previously determined in what time intervals it will transmit and in what time intervals it will remain in a waiting state. In particular, to determine at what time intervals it is to be transmitted and at what time intervals it is to remain in a waiting state, the node considered functions as described below.

In particular, when the considered node enters the network (for example, it activates or approaches another node of the network until it is at a distance from it that allows the transmission of subframes from one to another), it is kept waiting during N first intervals, in which it receives the possible subframes transmitted by other nodes, and stores the FI vectors contained therein. During said first N time intervals, the considered node does not delete any FI vector from its own matrix data structure.

Then, based on the stored FI vectors, the node considered determines whether there are accessible time intervals between the N future time intervals (which have not yet elapsed and, consequently, are subsequent to the first N time intervals in which the node has been waiting). In particular, a time interval z of future time intervals is accessible if all the fields of the stored FI vectors corresponding to the archetype range z mod N (ie column j = z mod N of the matrix data structure ) present the free status indication or the default indication.

Given the possible accessible time intervals, the considered node selects among them the chosen time interval. For example, if there is only one accessible time interval, the chosen time interval will coincide with the accessible time interval; instead, if there are several accessible time intervals, the chosen time interval will be one of the accessible time intervals chosen at random. Henceforth, it is assumed, without prejudice to the generality of the provisions, that the node considered will choose a lousy time interval p as the chosen time interval.

Once the chosen time interval has been selected, the considered node waits for the end of the time interval preceding the chosen time interval (the time interval p-1), while continuing to store the FI vectors of the possible subframes received, as previously described, and deleting, at the beginning of each time interval k, the stored vector FI (ie, the row indexed by i = k mod N) and associated with the archetype interval k mod N. At the end of the time interval p-1, the considered node verifies that the time interval p is really free; that is, it verifies that the fields of the stored FI vectors corresponding to the archetype time interval p mod N still contain the free status indication or the default indication. Hereinafter, it is assumed that the time interval p corresponds to the archetype time interval TX2 (p mod N = TX2).

If the chosen time interval is not really free, the node considered repeats the operations indicated above to determine new accessible time intervals and, subsequently, a new chosen time interval.

On the other hand, if the time interval p is really free, the node considered transmits the subframe during the time interval p, adding status indications to the FI vector fields of the transmitted subframe, as described above. In addition, the node considered deletes, from the matrix data structure, the vector FI associated with the archetype time interval p mod N (equivalent to the row indexed by i = p mod N).

After transmission of the subframe in the time interval p, the considered node receives, in the subsequent N-1 time slots, the possible subframes transmitted by other nodes, stores the Fl vectors contained therein and suppresses, at the beginning of each time interval k, the vector Fl stored and associated with the archetype time interval k mod N. Next, the node examines the N-1 fields of the stored vectors Fl corresponding to the archetype time interval TX2 and verifies that they contain the default indication, containing the free status indication or containing the busy status indication and the node's temporary source indicator.

If the above is fulfilled, the node considered interprets that the transmission was successful and can transmit a new subframe in the next time interval, that is, in the time interval p + N, which

it also corresponds to the archetype time interval TX2. To verify that the transmission of the new subframe was successful, the nodes considered repeat the operations described above, and so on.

On the contrary, that is, if one or more of the N-1 fields examined contain the indication of busy status and the temporary identifiers of origin of occupant nodes other than the node itself or in case they present the indication of collision , the node considered will take into account the priority identifier itself, that is, the priority identifier added to the transmitted subframe and the possible priority identifiers present in said one or more of the N-1 fields examined.

If the maximum priority identifier is the priority identifier itself, the considered node can transmit a new subframe in the next time interval p + N. Instead, if the maximum priority identifier is contained in one of the N-1 fields examined, the considered node estimates that the time interval is occupied by the node indicated by the temporary origin identifier contained in the examined field that contains The highest priority identifier. In concession, the node considered refrains from transmitting and repeats the operations described previously, beginning with the determination of the possible accessible time intervals.

In the particular case that there are two or more maximum priority identifiers among the priority identifiers considered, and one of those maximum priority identifiers is the priority identifier itself, the node considered will attempt to transmit a new subframe in the following interval of p + N time, will be kept waiting for subsequent N-1 subframes and will repeat the verifications described above. If it is verified again that there are two or more maximum priority identifiers, among which the priority identifier itself is present, the node considered transmits, with a first probability, a new subframe in the time interval p + 2N, and refrains of transmitting, with a second probability. In the latter case, the node considered determines new accessible time intervals, chooses a new time interval chosen and continues in the manner described above.

According to an embodiment of the present process, when determining the time interval chosen at the end of the first N time intervals mentioned above and in the absence of accessible time intervals, the considered node can take into account the N intervals of time following the first N time intervals and transmitting, in the time interval corresponding to the archetypal time interval, the corresponding FI vector fields stored by the considered node containing the busy status indication and the minimum priority identifier, to thereby guarantee the quality of service for high priority traffic.

It must be taken into account on the other hand that, at the time of transmitting the first subframe after entering the network, the node considered presents N FI vectors in memory. In other words, at that time, the matrix data structure implemented by the node considered has the dimensions NxN. Once the first subframe has been transmitted, at each instant the node considered has N-1 Fl vectors in memory, as described above.

According to the present procedure, it is therefore possible to distinguish between free and busy time intervals and time intervals in which a collision has occurred, propagate the collision information and, consequently, prevent network failures. In particular, in comparison to the RR-AIoha protocol, not only is the determination, for each of the N archetypal time intervals of the archetype frame, a corresponding status indication chosen from four different status indications, but All N status indications determined by each node are also propagated by the node itself, through the transmission of its own sub-frames, to the other nodes of the network. In this way, this procedure guarantees a high quality of service, even in the presence of rapidly evolving environments.

Again with reference to the example described in related figure 2, upon receipt of the subframe transmitted by node n6 in time interval five, nodes n1, n2 and n3 detect a collision in time interval two, since in the interval of time two the transmission of the node n2 has been detected, while the vector Fl contained in the subframe transmitted by the node n6 indicates that the interval of time two is occupied by the node n4. Therefore, according to the present procedure, nodes n1, n2 and n3 will transmit subframes with vectors Fl, in which the fields corresponding to the archetype time interval two have the busy bit equal to "0" and the CLS bit equal to "1 ", which will also allow the other nodes to detect the collision.

Likewise, according to the present procedure, the possibility of transmitting indirect connection indications makes it possible to prevent the possibility of propagating information associated with the transmission by a specific node, to other nodes of the network, over more than two jumps.

In this sense, without prejudice to the generality of the above, it is considered that the network represented in Figure 5 is constituted by the nodes na, nb, nc and nd. In particular, nodes na and nb are directly connected to each other, that is, node nb can receive subframes transmitted by node na, and vice versa. Likewise, nodes nb and

nc, as well as the nodes nc and nd, are also directly connected. In contrast, node nc cannot receive subframes transmitted by node na, for example due to excessive distance, and node nd cannot receive subframes transmitted by node na or subframes transmitted by node nb. In concession, in technical jargon it is said that node nb is at a distance of one jump from node na, it is said that node nc is at a distance of two jumps from node na, that is, it is indirectly connected to the node na by interposition of an intermediate node (node nb), and it is said that node nc is at a distance of three hops from node na, that is, it is connected to node na by interposition of two intermediate nodes (nodes nb and nc). It is also assumed that the number N of time intervals per frame is equal (for example) to ten, and that both the time intervals and the archetype time intervals are numbered starting with the number one. Finally, it is assumed that the nodes na, nb, nc and nd transmit, respectively, in time interval two, time interval five, time interval nine and time interval ten.

According to the present procedure, the node na transmits in the interval of time two a first subframe, the vector Fl of which contains a field corresponding to the archetype time interval two, the busy bit and the CLS bit present in said field respectively being equal to "1" and "0". Node nb receives said first subframe; therefore, node nb considers that time interval twelve is reserved, then does not transmit in it.

Then, when the node nb transmits a second subframe in time interval five, the second subframe contains a vector Fl in which the field corresponding to archetype time interval two still has the busy bit equal to "1" and the bit CLS equal to "0". The second subframe is received not only by node na, but also by node nc, which therefore considers that time interval twelve is reserved.

Then, the node nc transmits, in the interval nine for example, a third subframe in which, among other things, the field of the vector FI corresponding to the archetype time interval two presents the busy bit equal to "1" and the bit CLS equal to "1". Therefore, upon receipt of the third subframe, node nd considers that time interval twelve is reserved; however, at the time of transmitting a fourth subframe, in time interval ten for example, said node sets the busy bit and the CLS bit of the field corresponding to time interval two archetype to "0". In this way, any possible additional node (not shown) connected to node nd will consider that the time interval twelve is accessible and may transmit subframes therein, without causing any collision.

Therefore, the present procedure allows the reuse of a time interval, in which a certain node has made a transmission, by nodes that are at a distance of more than three hops from the node that has made the transmission. In this way, the performance levels of the communications network are optimized thanks to the reuse of time intervals. In addition, the present procedure allows to overcome the so-called "hidden terminal" problem; that is, it prevents the generation of any inconsistency between the status indications stored by the different nodes that may occur, for example, in situations where a given node is visible to an additional node, but not to other nodes in the network .

According to another embodiment, when the node considered enters the network after waiting for N time intervals, the value of the originating time identifier itself is determined by choosing any value comprised (for example) between zero and two hundred and fifty-five, provided that This is not being used by other nodes of the previous N. In case there is no value not used by other nodes, the node considered randomly chooses a value between zero and two hundred fifty-five. Hereinafter, the value chosen as the temporary origin identifier by the node considered at the time of the first transmission will be called the "first label".

In particular, when the node considered transmits a subframe for the first time, for example in a transmission time interval corresponding to the archetype time interval TX3, the busy status indication and the first label are added to the corresponding vector field Fl to the archetype time interval TX3.

At the time of receipt, the other nodes that receive the subframe store, instead of the first tag, a second tag, obtained for example by calculating a function whose arguments are the first tag and the MAC address of the node considered, being present said MAC address in the payload portion of the subframe transmitted by the considered node. For example, this function can calculate the hash value of its own arguments. Then, when the other nodes in turn transmit the subframes, instead of the first tag, the second tag is added to the field of vector Fl corresponding to the archetype time interval TX3.

Next, to verify that the transmission itself was successful, the node considered verifies that the N-1 fields corresponding to the TX3 archetype time interval of the N-1 Fl vectors stored in the N-1 time intervals after the transmission understand, if they contain the busy status indication, the second label mentioned above. Said second tag is, in fact, known by the node considered, to the extent that it can be calculated based on the first tag and the MAC address itself.

In a subsequent transmission, the node considered uses a third tag, obtained (for example) by calculating the function mentioned above and using the second tag and the MAC address itself as arguments of the function. At the time of receipt, the other nodes in turn calculate a fourth tag, and so on.

In this way, also in case two nodes adopt, in the respective first transmissions (in one

5 single and same frame), a single and same first tag, it is likely that the respective second tags will be different to the extent that they will be calculated with different arguments (MAC addresses), thereby allowing the detection of possible collisions that affect The two nodes. If the second tags are also the same, it is likely that the respective third tags are different in any case, and so on. In addition, it is also possible to detect collisions using labels consisting of a limited number of bits, by

For example, less than eight bits, thereby limiting the amount of additional information present in the subframes.

It should also be noted that the fact that the nodes do not keep in memory more than N Fl vectors means that the use, for each node, of different tags in different transmissions does not entail any

15 ambiguity Actually, it is not necessary that the nodes indicate precisely nodes located at a distance of more than one jump, but that it is sufficient that they detect the possible occupation of the time intervals or the presence of possible collisions.

The advantages offered by this access procedure are clearly shown in the description 20 above.

Finally, as will be apparent, it is possible to make modifications and variants of the access procedure described without departing from the scope of the present invention.

25 For example, the nodes may not store exactly the FI vectors present in the received subframes, but instead process the status indications, possibly resorting to additional indicators. For example, in the case of a field of an FI vector received in a reception time interval and containing the indirect connection indication, it is possible to store the free state indication, storing in a suitable indicator the impossibility of transmitting in a next time interval corresponding to the interval

30 archetype time corresponding to the reception time interval. Again, it is not necessary for deletion operations of stored FI vectors to be performed at the beginning of the time interval. Finally, it is possible to apply the present procedure together with other protocols of a known type, for example alternating periods of time in which the present access procedure is applied, to other periods of time in which a protocol of known type is applied.

Claims (12)

1. Procedure for accessing a communication channel for a communications network comprising a 5 plurality of nodes configured to transmit and receive subframes through said communication channel, said procedure comprising the following steps:

-

 define a sequence of frames, each frame consisting of one and the same first number (N) of time intervals, and each time interval being configured to accommodate the transmission of a single subframe and

10 being associated with a corresponding period of a reference frame formed by a number of periods equal to said first number (N) of time intervals; Y

-

 selectively execute, at each time interval and for each particular node of the communications network,

one of the following operations: 15

-

 determine, for each period of the reference frame, a corresponding status indication, and transmit a proper subframe, said transmission stage comprising in said own subframe a vector (FI) formed by a number of fields equal to said number of periods , each field referring to a corresponding period, and entering in each of said fields the status indication determined for the corresponding period; Y

-

 receiving a possible subframe transmitted by another node and storing information regarding the status indications contained in the vector (Fl) included in said received subframe;

said access procedure being characterized in that said step of determining a corresponding Status indication for each period of the reference frame comprises selecting by the determined node, alternatively, a free status indication, a busy status indication, a collision indication or an indirect connection indication, which are all different from yes; and wherein said step of storing information regarding the status indications contained in the vector (F) included in said received subframe comprises storing the vector (FI) included in said received subframe and associating said vector (FI) 30 included in said subframe received for the period corresponding to the time interval of reception by the determined node of said subframe received; said procedure further comprising carrying out, by the determined node, in case no subframe has been transmitted or received in a time interval, the storage stage of a default vector, the fields of which contain a default indication ; and in which said stage of transmission of an own subframe comprises, for each field considered between the fields 35 of the vector (FI) included in said own subframe, if the status indication entered in the field considered is the indication of busy status, the collision indication or the indirect connection indication, also enter a source identifier in the field considered, which identifies a respective occupying node; and wherein said selection stage by the node determined, alternatively, of a free status indication, an occupied status indication, a collision indication or an indirect connection indication comprises, in a 40 first interval:

-

 select, for the period corresponding to the first time interval, the indication of busy status;

-

 select each of the remaining periods of the reference frame and, for each selected period,

45 select the fields of stored vectors (Fl) corresponding to the selected period and then select, for the selected period:

-

 the free status indication, if the selected fields contain the default indication or the indication of

free state, or if at least one of the selected fields contains the indirect connection indication 50 and the other selected fields contain the default indication, or the free status indication;

-

 the busy status indication, if a detection field that is part of the selected fields and is contained in the vector (Fl) stored and associated with the selected period contains the busy status indication and a first origin identifier, and the other fields from among those selected fields contain the indication

55 by default, either they contain the free status indication or the indirect connection indication, or they contain the busy status indication and said first origin identifier;

-

 the collision indication, if a first field and a second field among those selected fields contain

the indication of busy status and different origin identifiers and 60

-

 the indirect connection indication, if said detection field contains the default indication, and one or more of the other selected fields contain the busy status indication and the same origin identifiers.

2. Access method according to claim 1, further comprising performing, for each determined node 65 of the communications network, the following steps:

-

 access the communications network; Y

-

 wait for a second number of time intervals and subsequently delete, in each successive time interval, the vector (Fl) stored and associated with the period corresponding to said subsequent time interval.

3. Access method according to claim 2, further comprising performing, for each particular node of the communications network, the following steps:

-

 during said second number of time intervals, execute the reception and storage operation;

-

 after said second number of time slots, select, based on stored vectors (Fl), a number of accessible time slots from a number of time slots subsequent to said second number of time slots, said number being time intervals after said second number of time intervals equal to said first number (N) of time intervals;

-

between said number of accessible time intervals, select a chosen time interval, said chosen time interval corresponding to a chosen period;

-

 execute the reception and storage operation up to the time interval preceding the chosen time interval, and subsequently verify a first criterion comprising verifying if the fields of the vectors (Fl) stored and corresponding to the chosen period contain, alternatively, the indication Free status or default indication; Y

-

 if said first criterion is met, transmit in the chosen time interval; otherwise, determine a number of new accessible time intervals and select a new chosen time interval.

Four.

 Access method according to claim 3, wherein said second number of time slots is equal to said first number (N) of time slots.

5.

 Access method according to claim 3 or claim 4, wherein said step of selecting a number of accessible time slots comprises selecting each time slot from among said number of time slots after said second number of time slots. , and verify that said time interval meets a second criterion comprising verifying that the fields of the vectors (FI) stored and corresponding to the period corresponding to the selected time interval, contain, alternatively, the free status indication or the indication by default.

6.

 Access method according to claim 5, wherein the transmission stage of a subframe itself is performed in a transmission time interval; the procedure also comprising the realization, for each particular node of the communications network, of the following steps:

-

 after a number of additional time intervals equal to said first number (N) of time intervals minus one, verify that the fields of the stored vectors (FI) and corresponding to the period corresponding to the transmission time interval, contain the default indication, or have the indication of free status or contain the indication of busy status and a source identifier indicating the given node.

7.

 Access method according to claim 6, wherein said step of transmitting an own subframe comprises introducing a priority identifier in each field of the vector (FI) included in said own subframe.

8.

 Access method according to claim 7, wherein said step of selecting a number of accessible time intervals comprises, in case no time interval selected from said number of time intervals after said second number of time intervals time meets said second criterion, select a single time interval accessible from said number of time intervals after said second number of time intervals, according to the priority identifiers entered in fields of stored vectors (FI).

9.

 Access method according to claim 8, wherein said step of introducing an origin identifier comprises, for each transmitted subframe, the following steps:

-

 enter in the field of the vector (FI) included in the single transmitted subframe and corresponding to the period corresponding to the transmission time interval of the single subframe, a respective first label associated with the determined node and said single transmitted subframe;

-

 enter in the other fields of the vector (FI) included in the only transmitted subframe, respective second labels, which are functions of the first labels contained in the subframes previously received by the

determined node and the information contained in said previously received subframes. 10. Node for a mobile communications network, configured to implement all stages of the access procedure according to any of the preceding claims. 5

eleven.

 Mobile communication network comprising at least one node according to claim 10.

12.

Mobile communication network according to claim 11, characterized in that it is of the wireless type.

10 13. Software product that can be loaded into a memory of a node of a communications network and is configured to implement, when executed, all the steps of the access procedure according to any one of claims 1 to 9.